Skip to content

ROADMAP.md

Latest commit

 

History

History
85 lines (65 loc) · 4.01 KB

ROADMAP.md

File metadata and controls

85 lines (65 loc) · 4.01 KB
Raw

🗺️ Learning Roadmap

The complete path from zero to production machine learning. Chapters are ordered so each builds on the last, but they are self-contained enough to jump around once you have the foundations (01–08).

Legend: ✅ authored · 🧩 scaffolded (structure + outline ready to fill)

Phase 1 — Foundations

# Chapter You will learn Status
01 Introduction What ML is, AI vs ML vs DL, types of learning, the ML workflow ✅
02 Python Review Variables, functions, OOP, generators, decorators, typing, NumPy, Pandas 🧩
03 Linear Algebra Vectors, matrices, dot products, eigenvalues/vectors, SVD 🧩
04 Statistics Descriptive stats, distributions, expectation, variance, sampling 🧩
05 Probability Probability rules, Bayes' theorem, MLE, entropy, information theory 🧩

Phase 2 — Data Skills

# Chapter You will learn Status
06 Data Preprocessing Missing values, encoding, scaling, normalization, pipelines 🧩
07 Data Visualization Matplotlib, Plotly, EDA, correlation, distributions 🧩
08 Feature Engineering Creation, selection, transformation, leakage avoidance 🧩

Phase 3 — Classical Machine Learning

# Chapter You will learn Status
09 Supervised Learning The supervised framing, train/test, bias–variance 🧩
10 Regression Linear & polynomial regression, gradient descent, regularization ✅
11 Classification Logistic regression, KNN, Naive Bayes, SVM, decision trees 🧩
12 Unsupervised Learning The unsupervised framing, similarity, distance metrics 🧩
13 Clustering K-Means, DBSCAN, hierarchical clustering 🧩
14 Dimensionality Reduction PCA, t-SNE, UMAP 🧩

Phase 4 — Model Craft

# Chapter You will learn Status
15 Model Evaluation Cross-validation, metrics, ROC/PR curves, SHAP 🧩
16 Ensemble Learning Bagging, boosting, Random Forest, XGBoost, LightGBM, CatBoost 🧩
17 Anomaly Detection Isolation Forest, one-class SVM, statistical methods 🧩
18 Time Series Stationarity, ARIMA, Prophet, LSTM forecasting 🧩
19 Recommendation Systems Collaborative & content filtering, matrix factorization 🧩

Phase 5 — Deep Learning

# Chapter You will learn Status
20 Intro to Deep Learning Perceptron, activation functions, forward pass 🧩
21 Neural Networks Backpropagation, optimizers, loss functions (from scratch) 🧩
22 CNN Convolutions, pooling, image classification 🧩
23 RNN Sequences, LSTM, GRU 🧩
24 Transformers Attention, self-attention, encoder/decoder 🧩
25 LLMs BERT, GPT, embeddings, fine-tuning, RAG, agents 🧩
26 Reinforcement Learning MDPs, Q-learning, policy gradients 🧩

Phase 6 — Production

# Chapter You will learn Status
27 MLOps MLflow, DVC, experiment tracking, model registry, drift detection 🧩
28 Model Deployment FastAPI, Docker, Streamlit, Gradio, ONNX 🧩
29 Cloud ML Deploying to Render, Railway, AWS, Azure, GCP 🧩

Phase 7 — Capstones

# Chapter You will learn Status
30 Capstone Projects 18+ end-to-end projects combining everything above 🧩

How to use this roadmap

  • Complete beginner? Go in order, 01 → 30.
  • Know Python already? Skim 02, then start at 03.
  • Have a CS/math background? Jump to Phase 3 and use 03–05 as reference.
  • Preparing for interviews? Do each chapter's Interview Questions section plus the interview/ bank.

Each chapter follows the same 14-part structure defined in book/TEMPLATE_CHAPTER.md.